0

enter image description here

I am using DP83620SQ/NOPB PHY for our 10/100Mbps Copper Media Ethernet Interface. We had connected through the Magnetics part 749013011A which is not PoE supported. But we need to use this for powering external devices without following the Standard PoE process (simply appending power with the data signals).

For PoE interface, the legacy (general) bob smith termination would not suit. Hence the 10nF capacitor is added along with the 75Ω termination resistor. To increase the power handling capacity through the bob smith termination during surge condition, we are purposely added one more 10nF+75Ω termination resistor combination along with the existing each 10nF+75Ω termination combination.

Will this scheme affect the cable characteristic impedance matching scheme?

zx485
  • 230
  • 4
  • 9
Dev
  • 1
  • 2
  • 1
    Draw a circuit and use fewer words. – Andy aka Aug 31 '16 at 11:34
  • Also, to your request of urgency: this is a Q&A site, not a (paid) contractor doing a design review for you/your company. – user2943160 Aug 31 '16 at 12:08
  • Thanks for the comment... Since I copied the mail content whatever I sent to vendor through mail, the strong line "Since we are closer to the LAYOUT release, expecting reply at very earliest." is pasted here.. – Dev Aug 31 '16 at 12:30

1 Answers1

1

The capacitors in series with 75Rs are for blocking the DC power signal to leak thru the 75Rs.

There is no need to parallel the RCs twice, instead connect two RCs to the loaded pairs (A, B: pins 1 & 2, 3 & 6) and other two RCs to the spare pairs (C, D: pins 4 & 5, 7 & 8). The 10nF's will not save from surge/ESD/fast transients itself, the 1nF/2kV is intended to do this.

In this question it is shown that in modern schemes BS termination has low effect.

Community
  • 1
user126247
  • 11
  • 1